Election pollsters 'got samples wrong'

Samantha and David Cameron after the Conservatives' election victory

The downfall of pollsters to forecast the outcome of the general election was largely due to “unrepresentative” count samples, an inquiry has found.

The polling industry came under enthusiasm for predicting a virtual dead heat when the Conservatives ultimately obtained on to outpoll Labour by 36.9% to 30.4%.

A nel of experts has concluded this was due to Tory voters being under-represented in phone and online asks.

But it said it was impossible to say whether “late swing” was also a factor.

The more than half of polls taken during last year’s five-week election effort suggested that David Cameron’s Conservatives and Ed Miliband’s Labour were neck-and-neck.

This led to wondering that Labour could be the largest rty in a hung rliament and could potentially be experiencing to rely on SNP support to govern.

But, as it turned out, the Conservatives secured an overall the better in May for the first time since 1992, winning 99 more behinds than Labour, their margin of victory taking nearly all commentators by their heels.

‘Statistical consensus’

The result prompted the polling industry to launch an aside from inquiry into the accuracy of their research, the reasons for any inaccuracies and how ballots were analysed and reported.

An interim report by the nel of academics and statisticians build that the way in which people were recruited to take rt – praying about their likely voting intentions – had resulted in “systematic over-representation of Labourers voters and under-representation of Conservative voters”.

These oversights, it found, had dnouement developed in a “statistical consensus”.

How opinion polls work

Replica caption The exit poll conducted on election day itself came much closer to the fundamental result than any of those conducted in the run-up

Most general nomination opinion polls are either carried out over the phone or on the internet. They are not unreservedly random – the com nies attempt to get a representative sample of the population, in age and gender, and the text is adjusted afterwards to try and iron out any bias, taking into account preceding voting behaviour and other factors.

But they are finding it increasingly onerous to reach a broad enough range of people. It is not a question of size – larger sample sizes are not necessarily more accurate.

YouGov, which y backs a nel of thousands of online volunteers to complete surveys, admitted they did not acquire access to enough people in their seventies and older, who were various likely to vote Conservative. They have vowed to change their methods.

Phone polls have good coverage of the population, but they suffer from low return rates – people refusing to take rt in their surveys, which can standard to bias.

How pre-election polls got it wrong

General Election poll tracker

This, it bring to light, was borne out by polls taken after the general election by the British Voting Study and the British Social Attitudes Survey, which produced a much varied accurate assessment of the Conservatives’ lead over Labour.

NatCen, who controlled the British Social Attitudes Survey, has described making “repeated strains” to contact those it had selected to interview – and among those most away reached, Labour had a six-point lead.

However, among the harder-to-contact com ny, who took between three and six calls to track down, the Conservatives were 11 points winning.


Former Liberal Democrat leader Lord Ashdown, who capitally dismissed the election night exit poll with a promise to eat his hat if it was valid, said the polls had a “considerable” im ct on the way people voted.

He told BBC Transmit 4’s the World At One: “The effect of the polls was to hugely increase the power of the Rightist message and hugely decrease the power of the Liberal Democrat message, which was you prerequisite us to make sure they don’t do bad things.”

The peer claimed the “mood of the political entity” was for another coalition and voters were “surprised” when the Conservatives won thoroughly.

“I think, therefore, there is an argument to be made that it actually l bly altered the outcome of the election.”

Evidence of a last-minute swing to the Conservatives was “inconsistent”, the winning experts said, and if it did happen its effect was likely to have been ordinary.

They also downplayed other potential explanations for why the polls got it flop, such as misreporting of voter turnout, problems with question phrasing or how overseas, postal or unregistered voters were treated in the polls.

In any case, the nel said it could not rule out the possibility of “herding” – where settle downs configured their polls in a way that caused them to deviate illiberal than could have been expected from others understood the sample sizes. But it stressed that did not imply malpractice on behalf of the firms perturbed.

Prof trick Sturgis, director of the National Centre for Research Methods at the University of Southampton and manage of the nel, told the BBC: “They don’t collect samples in the way the Office for Citizen Statistics does by taking random samples and keeping knocking on doors until they pre re got enough people.

“What they do is get anyone they can and try and match them to the residents… That approach is perfectly fine in many cases but from time to time it goes wrong.”

Prof Sturgis said that sort of allocation sampling was cheaper and quicker than the random sampling done by the equals of the ONS, but even if more money was spent – and all of the inquiry’s recommendations were all implemented – samples would still never get it right every time.

Analysis by the BBC’s public editor Laura Kuenssberg


I remember the audible catch ones breath in the BBC’s election studio when David Dimbleby read out the exit ballot results.

But for all that the consequences of that startling result were sundry and various, the reasons appear remarkably simple.

Pollsters didn’t ask reasonably of the right people how they planned to vote. Proportionately they implored too many likely Labour voters, and not enough likely Conservatives

Political science is not a precise science and predicting how people will vote will peacefulness be a worthwhile endeavour. Political rties, journalists, and the public of course desire be foolish to ignore them. But the memories and embarrassment for the polling industry of 2015 transfer take time to fade.

Read Laura’s blog on how the pundits and pollsters got it out of order

Joe Twyman, from pollster YouGov, told the BBC it was becoming increasingly ticklish to recruit people to take rt in surveys – despite, in YouGov’s prove, ying them to do so – but all efforts would be made to recruit subjects in “a uncountable targeted manner”.

“So more young people people who are disengaged with diplomacy, for example, and more older people. We do have them on the nel, but we have occasion for to work harder to make sure they’re represented sufficiently because it’s absolved they weren’t at the election,” he said.

On Tuesday, the Lords approved a restaurant check which would create a new watchdog to regulate future polling by denominating sampling methods, producing guidance on the wording of questions and deciding whether there should be a suspension on polls in the run-up to elections.

But the private member’s bill, tabled by Laboriousness’s Lord Foulkes, has yet to be introduced to the Commons and is unlikely to become law due to lack of orderly time.

Leave a Reply

Your email address will not be published. Required fields are marked *